According to HTTP spec, the grammar for defining the charset is

charset = token

Some http stacks (e.g. Sonic) may wrap the token in double quotes, e.g.
charset="utf-8"

rather than
charset=utf-8

The presence of these quotes make Camel raise an error right now
java.io.UnsupportedEncodingException: "utf-8"

Although Camels current behavior is correct, we should make Camel more fault tolerant and make it process charset tokens wrapped in quotes.